Controllable Swarm Animation Using Deep Reinforcement Learning With a Rule-Based Action Generator.
Zong-Sheng WangChang Geun SongJung LeeJong-Hyun KimSun-Jeong KimPublished in: IEEE Access (2022)
Keyphrases
- reinforcement learning
- action selection
- action space
- partially observable domains
- function approximation
- temporal difference
- particle swarm optimization
- reward shaping
- state space
- state action
- swarm intelligence
- reinforcement learning algorithms
- transition model
- user interface
- cooperative
- rule base
- agent learns
- collective behavior
- rule based systems
- model free
- data driven
- expert systems
- learning algorithm
- virtual environment
- computer animation
- policy search
- markov decision process
- computer graphics
- mobile robot
- search space
- fitted q iteration